Repository navigation
feat: route artifact writing to a writer agent and add Recommendations - #32
Merged
Merged
Conversation
The Strict output style carried a second writing layer for durable artifacts, derived from ASD-STE100 Simplified Technical English. Remove that layer and its cross-reference so the style governs conversational voice only. Artifact voice is handled elsewhere, not inside a swappable style.
Add an Artifact Authoring rule to the shared workflow guidelines so all runtimes (Claude, Codex, OpenCode) route durable artifact writing to a writer subagent matched to the artifact type, rather than writing the artifact in the coordinating agent. This replaces the removed STE-based artifact voice with a dispatch rule that lives outside any single runtime or swappable output style.
Introduce a writer subagent that rewrites finished technical material into readable engineering documentation. The shared body lives in a gotemplate, with thin Claude and OpenCode bindings that pin the mid-tier model for each runtime. Point the Artifact Authoring rule at this agent as the concrete example of routing durable artifact writing to a voice filter.
Give advisory content a dedicated home so it stops landing in Findings. Recommendations sits between Done and Next, holds advice the reader can accept or decline, and keeps required decisions whole in Next. Update the section contract, the escape hatch, the self-check, and the UserPromptSubmit reinforcement hook so all four agree on five sections.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
📝 Summary
The Strict output style carried a second, STE-derived writing layer for durable artifacts that made prose stiff and clipped. This branch removes that layer, moves durable-artifact writing to a dedicated writer subagent shared across runtimes, and adds a
Recommendationssection to the Strict style so advice stops landing inFindings.Key Decisions
🗑️ Remove the STE artifact mode
Dropped the ASD-STE100 Simplified Technical English layer from the Strict output style, along with its cross-reference. The style now governs conversational voice only; artifact voice is owned elsewhere.
🧭 Route artifact writing to a writer agent
Added an
Artifact Authoringrule to the shared workflow guidelines, so Claude, Codex, and OpenCode all route durable artifact writing to a writer subagent matched to the artifact type instead of writing it in the coordinating agent.🤖 mid-tier-technical-editor agent
Introduced the writer agent as a shared gotemplate body with thin per-runtime bindings. The
mid-tiername prefix follows the tier convention and pins Sonnet on Claude andopenai/gpt-5.6-terraon OpenCode. Codex is excluded for now.✋ Recommendations section
Added a fifth Strict section between
DoneandNext.Recommendationsis advisory (which option is favored and why, optional follow-ups);Nextstays required actions and decisions. Updated the section contract, escape hatch, self-check, and the UserPromptSubmit reinforcement hook so all agree on five sections.🔄 Before & After
mid-tier-technical-editor(Claude + OpenCode)